EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

827. Slowly varying regression under sparsity

Invited abstract in session TC-27: Mathematical Optimization for Trustworthy Machine Learning, stream Mathematical Optimization for XAI.

Tuesday, 12:30-14:00
Room: 047 (building: 208)

Authors (first author is the speaker)

1. Vassilis Digalakis
Information systems and operations management, HEC Paris

Abstract

I will present the framework of slowly varying regression under sparsity, allowing sparse regression models to exhibit slow and sparse variations, through an application in energy consumption prediction. First, I will formulate the problem of parameter estimation as a mixed-integer optimization problem; then, I will demonstrate that it can be precisely reformulated as a binary convex optimization problem through a novel relaxation technique, convexifying the non-convex objective function while matching the original objective on all feasible binary points. I will develop a highly optimized implementation of a cutting plant-type algorithm, a fast regularization-based heuristic method that guarantees a feasible solution, and a practical hyperparamrter tuning procedure relying on binary search that, under certain assumptions, is guaranteed to recover the true model parameters.

Keywords

Status: accepted


Back to the list of papers